NetBSD Problem Report #47342

From dholland@macaran.localdomain  Sun Dec 16 21:17:38 2012
Return-Path: <dholland@macaran.localdomain>
Received: from mail.netbsd.org (mail.netbsd.org [149.20.53.66])
	by www.NetBSD.org (Postfix) with ESMTP id A5DE563E529
	for <gnats-bugs@gnats.NetBSD.org>; Sun, 16 Dec 2012 21:17:37 +0000 (UTC)
Message-Id: <20121216211756.B2FEB6E232@macaran.localdomain>
Date: Sun, 16 Dec 2012 16:17:56 -0500 (EST)
From: dholland@eecs.harvard.edu
Reply-To: dholland@eecs.harvard.edu
To: gnats-bugs@gnats.NetBSD.org
Subject: devel/py-gobject3 build failure
X-Send-Pr-Version: 3.95

>Number:         47342
>Category:       pkg
>Synopsis:       devel/py-gobject3 build failure
>Confidential:   no
>Severity:       critical
>Priority:       medium
>Responsible:    pkg-manager
>State:          closed
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Sun Dec 16 21:20:00 +0000 2012
>Closed-Date:    Fri Sep 30 07:35:02 +0000 2016
>Last-Modified:  Fri Sep 30 07:35:02 +0000 2016
>Originator:     David A. Holland
>Release:        NetBSD 6.99.11 (20120906)
>Organization:
>Environment:
System: NetBSD macaran 6.99.11 NetBSD 6.99.11 (MACARAN) #15: Mon Oct 15 21:24:31 EDT 2012 dholland@macaran:/usr/src/sys/arch/amd64/compile/MACARAN amd64
Architecture: x86_64
Machine: amd64
>Description:

py-gobject3 croaks when building as follows:

In file included from /home/pkgbulk/c/obj/devel/py-gobject3/work/.buildlink/include/python2.7/Python.h:126:0,
                 from pygmaincontext.c:27:
/home/pkgbulk/c/obj/devel/py-gobject3/work/.buildlink/include/python2.7/modsupport.h:27:1: error: '_PyArg_ParseTuple_SizeT' is an unrecognized format function type
gmake[3]: *** [_glib_la-pygmaincontext.lo] Error 1

Digging around shows that this failure is here:

PyAPI_FUNC(int) PyArg_ParseTuple(PyObject *, const char *, ...) Py_FORMAT_PARSETUPLE(PyArg_ParseTuple, 2, 3);

and that the Py_FORMAT_PARSETUPLE business expands to

  __attribute__((format(PyArg_ParseTuple, 2, 3)))

and then to

  __attribute__((format(_PyArg_ParseTuple_SizeT, 2, 3)))

and apparently it expects gcc to understand this -- there are config
defines for whether to enable it, and they're turned on, even though
it evidently doesn't work and, I would think, can't seriously be
expected to work.

What is this about? And why is it breaking only in py-gobject3?

(Both python26 and python27 builds exhibit the same behavior.)

>How-To-Repeat:

make

>Fix:

>Release-Note:

>Audit-Trail:
From: Thomas Klausner <wiz@NetBSD.org>
To: NetBSD bugtracking <gnats-bugs@NetBSD.org>
Cc: 
Subject: Re: pkg/47342: devel/py-gobject3 build failure
Date: Sun, 16 Dec 2012 22:41:39 +0100

 On Sun, Dec 16, 2012 at 09:20:00PM +0000, dholland@eecs.harvard.edu wrote:
 > System: NetBSD macaran 6.99.11 NetBSD 6.99.11 (MACARAN) #15: Mon Oct 15 21:24:31 EDT 2012 dholland@macaran:/usr/src/sys/arch/amd64/compile/MACARAN amd64
 ...
 > py-gobject3 croaks when building as follows:

 Just for the record, py-gobject3 built from latest pkgsrc builds fine
 for me on 6.99.15/amd64.

 Let me know what kind of information could help you.
  Thomas

From: David Holland <dholland-pbugs@netbsd.org>
To: gnats-bugs@NetBSD.org
Cc: 
Subject: Re: pkg/47342: devel/py-gobject3 build failure
Date: Sun, 16 Dec 2012 22:12:25 +0000

 On Sun, Dec 16, 2012 at 09:45:04PM +0000, Thomas Klausner wrote:
  >  > System: NetBSD macaran 6.99.11 NetBSD 6.99.11 (MACARAN) #15: Mon Oct 15 21:24:31 EDT 2012 dholland@macaran:/usr/src/sys/arch/amd64/compile/MACARAN amd64
  >  ...
  >  > py-gobject3 croaks when building as follows:
  >  
  >  Just for the record, py-gobject3 built from latest pkgsrc builds fine
  >  for me on 6.99.15/amd64.
  >  
  >  Let me know what kind of information could help you.

 Hrm...

 Does $PREFIX/include/python2.7/pyconfig.h define
 HAVE_ATTRIBUTE_FORMAT_PARSETUPLE?

 -- 
 David A. Holland
 dholland@netbsd.org

From: Thomas Klausner <wiz@NetBSD.org>
To: NetBSD bugtracking <gnats-bugs@NetBSD.org>
Cc: 
Subject: Re: pkg/47342: devel/py-gobject3 build failure
Date: Sun, 16 Dec 2012 23:55:13 +0100

 On Sun, Dec 16, 2012 at 10:15:05PM +0000, David Holland wrote:
 >  Does $PREFIX/include/python2.7/pyconfig.h define
 >  HAVE_ATTRIBUTE_FORMAT_PARSETUPLE?

 No.

 /* Define if GCC supports __attribute__((format(PyArg_ParseTuple, 2, 3))) */
 /* #undef HAVE_ATTRIBUTE_FORMAT_PARSETUPLE */

  Thomas

From: David Holland <dholland-pbugs@netbsd.org>
To: gnats-bugs@NetBSD.org
Cc: 
Subject: Re: pkg/47342: devel/py-gobject3 build failure
Date: Sun, 16 Dec 2012 23:03:45 +0000

 On Sun, Dec 16, 2012 at 11:00:09PM +0000, Thomas Klausner wrote:
  >  On Sun, Dec 16, 2012 at 10:15:05PM +0000, David Holland wrote:
  >  >  Does $PREFIX/include/python2.7/pyconfig.h define
  >  >  HAVE_ATTRIBUTE_FORMAT_PARSETUPLE?
  >  
  >  No.
  >  
  >  /* Define if GCC supports __attribute__((format(PyArg_ParseTuple, 2, 3))) */
  >  /* #undef HAVE_ATTRIBUTE_FORMAT_PARSETUPLE */

 ok, let's see if I can figure out why I'm getting borked python packages.

 -- 
 David A. Holland
 dholland@netbsd.org

From: "David A. Holland" <dholland@netbsd.org>
To: gnats-bugs@gnats.NetBSD.org
Cc: 
Subject: PR/47342 CVS commit: pkgsrc/lang/python27
Date: Sun, 16 Dec 2012 23:22:29 +0000

 Module Name:	pkgsrc
 Committed By:	dholland
 Date:		Sun Dec 16 23:22:29 UTC 2012

 Modified Files:
 	pkgsrc/lang/python27: distinfo
 	pkgsrc/lang/python27/patches: patch-al

 Log Message:
 Use -Werror=format when checking HAVE_ATTRIBUTE_FORMAT_PARSETUPLE, so
 injecting -Wall -Wno-error does not cause the test to produce the wrong
 answer. (If it does, the wrong information is installed in /usr/include,
 and ultimately provokes PR 47342.)


 To generate a diff of this commit:
 cvs rdiff -u -r1.19 -r1.20 pkgsrc/lang/python27/distinfo
 cvs rdiff -u -r1.7 -r1.8 pkgsrc/lang/python27/patches/patch-al

 Please note that diffs are not public domain; they are subject to the
 copyright notices on the relevant files.

From: "David A. Holland" <dholland@netbsd.org>
To: gnats-bugs@gnats.NetBSD.org
Cc: 
Subject: PR/47342 CVS commit: pkgsrc/lang/python26
Date: Sun, 16 Dec 2012 23:32:46 +0000

 Module Name:	pkgsrc
 Committed By:	dholland
 Date:		Sun Dec 16 23:32:46 UTC 2012

 Modified Files:
 	pkgsrc/lang/python26: distinfo
 	pkgsrc/lang/python26/patches: patch-al

 Log Message:
 Use -Werror=format when checking HAVE_ATTRIBUTE_FORMAT_PARSETUPLE, so
 injecting -Wall -Wno-error does not cause the test to produce the wrong
 answer. (If it does, the wrong information is installed in /usr/include,
 and ultimately provokes PR 47342.)

 Same as -r1.8 of python27's patch-al.


 To generate a diff of this commit:
 cvs rdiff -u -r1.47 -r1.48 pkgsrc/lang/python26/distinfo
 cvs rdiff -u -r1.9 -r1.10 pkgsrc/lang/python26/patches/patch-al

 Please note that diffs are not public domain; they are subject to the
 copyright notices on the relevant files.

State-Changed-From-To: open->closed
State-Changed-By: dholland@NetBSD.org
State-Changed-When: Sun, 16 Dec 2012 23:58:30 +0000
State-Changed-Why:
fixed. (my fault)


From: Bernd Ernesti <netbsd@lists.veego.de>
To: gnats-bugs@NetBSD.org
Cc: 
Subject: Re: pkg/47342 (devel/py-gobject3 build failure)
Date: Mon, 17 Dec 2012 08:49:53 +0100

 On Sun, Dec 16, 2012 at 11:58:31PM +0000, dholland@NetBSD.org wrote:
 > Synopsis: devel/py-gobject3 build failure
 > 
 > State-Changed-From-To: open->closed
 > State-Changed-By: dholland@NetBSD.org
 > State-Changed-When: Sun, 16 Dec 2012 23:58:30 +0000
 > State-Changed-Why:
 > fixed. (my fault)

 Did I miss the bump of the packages version?

From: David Holland <dholland-pbugs@netbsd.org>
To: gnats-bugs@NetBSD.org
Cc: 
Subject: Re: pkg/47342 (devel/py-gobject3 build failure)
Date: Mon, 17 Dec 2012 13:51:35 +0000

 On Mon, Dec 17, 2012 at 07:55:03AM +0000, Bernd Ernesti wrote:
  >  On Sun, Dec 16, 2012 at 11:58:31PM +0000, dholland@NetBSD.org wrote:
  >  > Synopsis: devel/py-gobject3 build failure
  >  > 
  >  > State-Changed-From-To: open->closed
  >  > State-Changed-By: dholland@NetBSD.org
  >  > State-Changed-When: Sun, 16 Dec 2012 23:58:30 +0000
  >  > State-Changed-Why:
  >  > fixed. (my fault)
  >  
  >  Did I miss the bump of the packages version?

 Which packages? I didn't bump python because it only happens if you do
 non-default things; I didn't bump py-gobject3 because it didn't
 change.

 -- 
 David A. Holland
 dholland@netbsd.org

From: "David A. Holland" <dholland@netbsd.org>
To: gnats-bugs@gnats.NetBSD.org
Cc: 
Subject: PR/47342 CVS commit: pkgsrc/lang/python27
Date: Fri, 21 Dec 2012 20:26:42 +0000

 Module Name:	pkgsrc
 Committed By:	dholland
 Date:		Fri Dec 21 20:26:42 UTC 2012

 Modified Files:
 	pkgsrc/lang/python27: distinfo
 	pkgsrc/lang/python27/patches: patch-al

 Log Message:
 Revert previous; Joerg says it breaks on clang. (It would be nice if
 clang didn't masquerade as gcc and then turn out to be incompatible.)
 Reopens PR 47342, but I'll deal.


 To generate a diff of this commit:
 cvs rdiff -u -r1.20 -r1.21 pkgsrc/lang/python27/distinfo
 cvs rdiff -u -r1.8 -r1.9 pkgsrc/lang/python27/patches/patch-al

 Please note that diffs are not public domain; they are subject to the
 copyright notices on the relevant files.

From: "David A. Holland" <dholland@netbsd.org>
To: gnats-bugs@gnats.NetBSD.org
Cc: 
Subject: PR/47342 CVS commit: pkgsrc/lang/python26
Date: Fri, 21 Dec 2012 20:28:36 +0000

 Module Name:	pkgsrc
 Committed By:	dholland
 Date:		Fri Dec 21 20:28:36 UTC 2012

 Modified Files:
 	pkgsrc/lang/python26: distinfo
 	pkgsrc/lang/python26/patches: patch-al

 Log Message:
 Revert previous as it breaks clang. PR 47342


 To generate a diff of this commit:
 cvs rdiff -u -r1.48 -r1.49 pkgsrc/lang/python26/distinfo
 cvs rdiff -u -r1.10 -r1.11 pkgsrc/lang/python26/patches/patch-al

 Please note that diffs are not public domain; they are subject to the
 copyright notices on the relevant files.

State-Changed-From-To: closed->open
State-Changed-By: dholland@NetBSD.org
State-Changed-When: Fri, 21 Dec 2012 20:32:21 +0000
State-Changed-Why:
unfixed


State-Changed-From-To: open->closed
State-Changed-By: dholland@NetBSD.org
State-Changed-When: Fri, 30 Sep 2016 07:35:02 +0000
State-Changed-Why:
withdrawn by submitter


>Unformatted:

Home	PR Database Search